3 ms·
All sentences that start with "ChatGPT Fails to Understand" is correct because it's a language model, not a cogitation system.
by burnte 2y ago
All sentences that start with "ChatGPT Fails to Understand" is correct because it's a language model, not a cogitation system.
- goatlover 2y agoYet some think a cogitation system is just a language model. I don’t see how, since language is a late addition to brain evolution, and there’s more to understanding the world than predicting the next token.
- threatofrain 2y agoThat should be clear given that animals can explore causality but don't have formal language.
- burnte 2y agoI agree with you, life had significant cogitation hundreds of millions of years before language.
- Ukv 2y agoI think token generation may be a sufficiently general task to introduce pretty much any test of intellect - such as the causation question in this tweet. That does not imply the inverse, that cogitation requires language use.
- pelorat 2y agoIt's a memory retrieval and synthesis model, akin to human long term memory. I think it's a bit disingenuous to call true multimodal models for "language models". A prompt triggers a bunch of memories which gets recombined to satisfy the input. For cognition you need an agent that uses these recombined memories, a temporary scratch pad (short term memory) and an algorithm for solving the problem.
- wslh 2y agoThe interesting part of failures like this is adding it to the test suites, beyond the specific Turing Prize. As you could notice the author of this tweet knows something about computer science though.